The SIG trading interview is one of the most distinctive loops in quant finance. Susquehanna International Group built its process around a conviction the firm states openly: trading is decision-making under uncertainty, and the best way to find traders is to watch how people think in bets. That is why SIG, almost alone among the major prop shops, teaches new hires poker and probes it heavily in interviews. The loop tests the same skills as Jane Street or Optiver — mental math, expected value, probability — but filters hardest on whether you reason like a gambler in the good sense: sizing edges, folding bad hands, and staying rational when variance turns against you. This page covers the full process, what each round tests, the question types, the firm-specific nuances, and a multi-week prep plan.
The full process, end to end
A typical SIG quantitative trader pipeline runs like this:
- Online assessment / mental-math screen (~30–60 min). A rapid-fire arithmetic test — dozens of two-digit multiplications, percentages, and fractions against a steep clock — sometimes bundled with a personality / decision-making questionnaire and a sequence-pattern section. The cutoff is real; a weak math score ends the process here.
- First-round / phone interview (30–45 min). A conversation with a recruiter or junior trader: why trading, why SIG, and a few warm-up probability and EV questions. SIG uses it to gauge "think in bets" fit as much as math.
- Superday / onsite loop (4–6 rounds × 30–45 min). The core. You rotate through traders and formats: a probability and EV round, one or more poker / game-theory rounds, a market-making round, and a behavioral conversation. SIG runs many in person at its Bala Cynwyd, Pennsylvania headquarters, and the day is deliberately fatiguing.
- Poker / decision-making round. Distinctive enough to call out on its own. You may discuss pot odds and fold equity or work a betting scenario where the interviewer probes whether you size bets to your edge.
- Final / committee conversation. A senior trader panel that revisits probability, pressure-tests your reasoning, and covers offer and team fit (SIG has trading, quantitative research, and technology tracks).
The pipeline typically runs three to eight weeks for campus candidates, faster for experienced hires when a desk is waiting.
What the rounds actually test
SIG is explicit that it hires for a way of thinking, not a résumé. Across the loop, interviewers score four things:
- Edge recognition. Can you tell when a bet is favorable and by how much? Most rounds are disguised expected-value problems.
- Decision-making under uncertainty. Given incomplete information, do you make a rational choice and size it correctly — or freeze, over-hedge, or over-commit?
- Speed and accuracy in arithmetic. You cannot reason about a betting game in real time if three-digit multiplication slows you down. The mental-math bar is an entry ticket, not the finish line.
- Composure about variance. Traders watch how you respond when a bet goes wrong. Separate process from outcome — a good decision that loses is still a good decision — rather than tilting.
Question types by round
Mental-math screen
The arithmetic test is the first filter and is non-negotiable. Expect two- and three-digit multiplication, percentages of awkward numbers, fraction-to-decimal conversion, and division with remainders against a clock that forces a question every few seconds. It is in the same family as the Zetamac-style tests used across the industry, and the only preparation that works is volume — daily timed drilling until the operations are reflexive. Candidates who can do the math but not fast enough get filtered before any trader speaks to them.
Probability and expected value
The backbone of the loop. Problems are almost always framed as games you can bet on:
- EV of a game. "You pay a fixed amount to roll a die and receive the face value in dollars. What is the most you should pay?" Set it up cleanly and commit to a number.
- Sequential decisions. "Cards are turned over one at a time; you can stop whenever you like and take the current value. What is your stopping strategy?" These probe optimal stopping and conditional reasoning.
- Conditional probability. Bayesian updates framed as betting on which of two situations you are in, given some evidence.
- Coin and dice sequences. "How many flips on average until two heads in a row?" State-machine EV problems that reward the recurrence, not guessing.
The interviewer cares less about the exact number than whether your framing is sound and transparent.
Poker and game theory
SIG's signature round, and the one candidates least expect. The firm uses poker as a teaching tool because it isolates the core skill: profitable decisions with incomplete information, and managing risk across repeated bets. In the interview you might:
- Walk through a hand and explain whether to call, fold, or raise given pot odds and fold equity.
- Reason about bet sizing — how much to wager as a function of your edge and the information you reveal.
- Discuss bluffing frequency as a game-theoretic equilibrium rather than a gut call.
- Solve a simplified two-player betting game the interviewer invents on the spot.
You need not be a tournament player — just comfortable thinking probabilistically about edge, variance, and information.
Market-making and betting games
SIG also runs market-making-style rounds, weighted less heavily than at Jane Street or Optiver. The interviewer names a quantity with a hidden true value — windows on a building, the sum of two dice, the count of primes under 100 — and you quote a two-sided market (a bid and an ask) you will trade at. They trade against you, new information arrives, and you requote. It is the same edge-and-information reasoning as poker, in market language: quote tight enough to show confidence, wide enough to avoid being picked off, and update fast while tracking your position. The setup often doubles as a Fermi estimate, so the firm grades the honesty of your assumptions too.
Behavioral and fit
Shorter than in banking, but more pointed than at most trading firms because it ties directly to the "think in bets" thesis. Expect questions about a decision you made under uncertainty, a calculated risk, a bet that did not work out, and how you handle losing — probing whether you separate decision quality from outcome and stay rational under pressure.
SIG-specific nuances
A few things set the SIG loop apart and are worth preparing for explicitly:
- Poker is real, not a gimmick. Unlike firms where game-theory questions are occasional flavor, SIG organizes its trader culture and training around poker. Treat the poker round as a core technical round, and narrate your reasoning so they score the decision, not the result.
- The personality and decision questionnaire counts. SIG's assessment often includes situational-judgment and personality components alongside the math. Answer honestly and consistently; the firm is calibrating for a specific risk temperament.
- Bala Cynwyd, not Manhattan. SIG's headquarters and much of its trading floor sit outside Philadelphia, and onsites are frequently in person there. The culture skews collaborative and game-loving — interviewers respond well to candidates enjoying the puzzles.
- Multiple tracks. SIG hires trading, quantitative research, and technology roles with heavy early-round overlap. Confirm your track, since the later rounds and the coding bar diverge.
A multi-week preparation plan
Weeks 1–2 — Mental-math foundation. Daily timed arithmetic on two-digit multiplication, percentages, and fraction-to-decimal conversion until it is automatic. Target a top-tier Zetamac-style score before moving on, because the screen is a hard gate.
Weeks 3–4 — EV and probability. Drill expected-value problems, conditional probability, optimal stopping, and coin/dice sequences. Aim for 5–10 problems per day, framing each as a bet — "what is the EV, and how confident am I?" — until that framing is reflexive.
Weeks 5–6 — Poker, game theory, and market making. Learn pot odds, fold equity, and bet sizing well enough to reason out loud; if you can, play low-stakes poker with a partner. In parallel, practice market-making flows: quote, get traded against, update, requote, track P&L.
Week 7 — Estimation and behavioral. Drill Fermi problems with explicit assumption tracking. Build three to five "decision under uncertainty" stories, each framed around the bet you made, the information you had, and how you separated process from outcome.
Week 8 — Full mocks. Run full loops mixing mental math, EV, a poker scenario, and a market-making game in one sitting. SIG's superday is fatiguing; build the stamina to stay composed through the fifth round. The biggest differentiator is the same as at every trading firm — raw mental-math speed plus disciplined EV thinking — with an explicit poker layer on top.
How to practice for the SIG loop
The SIG bar is reasoning under live pressure, not silent solving, so the most effective preparation is repetition against an interviewer who pushes back. InterviewDen's quant-trading track is built for this loop. A voice-driven AI interviewer runs full rounds the way SIG does — market-making rounds where the AI takes the other side of your two-sided quotes and feeds you new information to requote on, EV and probability problems with live follow-ups, and mental-math drills against a timer you ratchet up as you improve. Because it is voice-first you practice reasoning out loud rather than typing numbers, and every session ends with a scored debrief flagging the composure issues SIG traders watch for — hesitation, price reversals, panic-widening, tilting after a bad trade. It is free to start.
For the conceptual foundation, work through the quant trading interview guide, which lays out the mental-math bar, the market-making flow, and the EV taxonomy, including how SIG weights game-theory and poker reasoning. To drill the question bank, quant trading brainteasers covers the EV, sequential-decision, and probability problems that fill out the SIG rounds. When you are ready to simulate the real thing, run a SIG-style trading mock — market-making plus EV rounds, mental-math drills, and a scored debrief in one session.
Common mistakes
- Treating poker as a throwaway. SIG weights it; reason about pot odds, fold equity, and bet sizing out loud.
- Optimizing for the outcome, not the decision. Defending a choice because it "worked out" is a red flag. Defend it on the information you had, and own a good decision that lost.
- Under-drilling mental math. The screen is a hard gate and the EV rounds assume arithmetic is free. Slow math caps your ceiling.
- Quoting markets too tight. A razor-thin spread to look confident gets picked off; wider-but-honest is the better signal.
- Over-explaining brainteasers. A clean setup in a minute beats a five-minute monologue.
- Tilting after a bad bet. Visible frustration is exactly what the firm screens against. Reset and quote the next problem cleanly.
- Not confirming your track. Trading, quant research, and technology diverge in the later rounds.
FAQ
Why does SIG ask about poker?
Because SIG sees poker as the closest everyday analog to trading: you decide with incomplete information, size bets to your edge, and stay rational across many hands as variance swings against you. The firm teaches poker to new traders and uses it in interviews to see whether you can reason about pot odds, fold equity, and bluffing as probabilistic problems. You do not need to be a strong player — you need to think in bets.
How hard is the SIG trading interview?
It sits in the top tier alongside Jane Street, Optiver, and IMC. The mental-math screen filters aggressively before you reach a trader, and the EV, poker, and market-making rounds reward genuine probabilistic fluency over memorized answers. Top trading firms accept single-digit percentages of final-round candidates.
What is the SIG mental-math test like?
A timed arithmetic screen — two- and three-digit multiplication, percentages, fractions, and division with remainders against a steep clock, in the same family as Zetamac-style tests, often bundled with a sequence-pattern section and a personality / decision questionnaire. The cutoff is real, so daily timed drilling to a top-tier score is the highest-leverage thing you can do early.
Do I need to be good at poker to get a SIG offer?
No. You need to reason fluently about the concepts poker exercises — pot odds, fold equity, bet sizing, bluffing frequency, decision-making under uncertainty — and be comfortable making a call you cannot verify. Talking through a hand matters far more than tournament results.
Does SIG ask coding questions?
It depends on the track. Pure trading seats are light on coding and heavy on probability, EV, and game theory. Quantitative research and technology roles carry a real coding bar — typically Python or C++ — so confirm your track and prepare accordingly.
What does SIG look for in the behavioral round?
Evidence that you think in bets. Expect questions about a decision you made under uncertainty, a calculated risk you took, and a bet that did not work out. SIG is checking whether you separate process from outcome, size risk rationally, and stay composed when things go against you.
How is the SIG interview different from Jane Street or Optiver?
All three test mental math, EV, and probability, but the emphasis differs. Optiver and IMC lean hardest on raw arithmetic speed; Jane Street emphasizes clean reasoning and market-making games. SIG is the game-theory and poker desk — it weights decision-making under uncertainty most heavily, and is the only one that builds its trader culture explicitly around poker.